A team of researchers at Rensselaer Polytechnic Institute is creating new artificial technology techniques that protect algorithms from vulnerabilities such as contaminated data, malicious attacks, or independent algorithm interference to improve patient care.
The researchers are being led by Assistant Professor of Biomedical Engineering Pingkun Yan and supported by a National Science Foundation Faculty Early Career Development Program (CAREER) Award.
“We see great potential in AI, machine learning, and deep learning in biomedical imaging,” said Yan in a press release. “We just need to build a system one step at a time to make it more robust, usable, and understandable.”
New Algorithm Could Improve Accuracy of Machine Learning Models
Artificial intelligence techniques show potential in advancing image reconstruction, image quality, computer-aid diagnosis, and image-guided surgery. However, according to Yan, deep learning adversaries remain an obstacle in the widespread application, resulting in confusing image results and inaccuracy.
“The adversary might cause the system to generate undesired or unwanted outputs,” Yan said. “The clinicians may be confused by the output of the system, or a wrong diagnosis may just slip through. It could cause significant medical errors in the diagnosis process and cause significant cost to our health care system.”
Currently, some tools can protect algorithms against adversaries. However, the tools will often decrease the quality of algorithmic performance. According to the press release, the goal of the research is to protect the algorithm by developing artificial intelligence techniques that can guard against deep learning glitches without diminishing the quality and performance of the algorithm.
The tool created by Yan and his team will be able to detect adversarial images, correct errors, and improve the quality of information produced by the AI system. However, Yan explained that new algorithms created by the team must consider the overall system, not just the individual parts,to ensure that accuracy is not sacrificed as efficiency increases.
“We envision such systems will help to make health care imaging more robust, more accurate, and will also boost the confidence of the users — especially the medical professionals,” Yan said
By protecting algorithms, artificial intelligence technology will better serve the public by producing accurate results and improving care.
(This is a slightly modified version of an article originally published in Health IT Analytics. The original article can be found at https://healthitanalytics.com/news/ai-technology-protects-algorithms-improves-patient-care)